Theor Appl Genet (1996J 93:703-709

9 Sprlnger-Verlag 1996

M. H. Bodanese-Zanettini 9 M . S . Lauxen S. N. C. Richter 9 S. Cavalli-Molina 9 C. E. Lange P. J. Wang 9 C. Y. Hu

Wide hybridization between Brazilian soybean cultivars and wild perennial relatives

Received: 2 January 1996 .' Accepted: 26 January 1996

soybean lines crossed with G 9943 were capable of obtained a greatly improved frequency of embryo rescue producing mature hybrid plants. There was no correlain intersubgeneric soybean hybrids. Successful crosses tion between the initial size of Group 2 seeds and plant were obtained in 31 different genotype combinations survival rate. The hybrids were cloned by grafting and between nine Brazilian soybean lines as the female treated with colchicine. One of the treated plants disparents and 12 accessions from Glycine canescens, G. played chromosome doubling. microphylla, G. tabacina and G. tomentella. The hybrid pod retention rate dropped to about 10% during the Key words Embryo rescue 9 Glycine spp. 9 first 8 days after pollination and stayed largely un- Intersubgeneric hybrids 9 Soybean 9 Wide changed up to the 20th day. Immature harvested seeds hybridization fell into three size groups: Group 1, smaller than 1.3 m m (mostly empty seed coats): Group 2, 1.9-5.0 ram; Group 3, larger than 5 m m (from selfing). A total of 90 putative Introduction hybrid embryos were rescued using a highly enriched B5 medium to nourish the newly dissected embryos. The Brazil is the world's second largest soybean-producing growing embryos were then placed in a high osmotic, and -exporting nation. Although participants in the modified B5 medium to induce maturation and dor- Brazilian soybean breeding program are aware of the mancy. Schenk and Hildebrandt medium was used to ever-changing needs of agriculture, the improvement germinate the dormant, partially dehydrated, physio- program is severely handicapped due to the cultivated logically mature embryos. Approximately 37~ of the Brazilian soybean gene pool being extremely limited rescued embryos developed into plantlets in vitro, and (Hiromoto and Vello 1986; Abdelnoor et al. 1995). Such approximately 8% grew into mature plants in the green- breeding difficulties are further compounded by the house. Morphological, cytological and isoenzyme pat- semitropical Brazilian climate; cultivated soybeans were terns confirmed the hybrid status of all seven mature originally introduced from countries with temperate plants, all of which were generated using G. tomentella climate. G 9943 as the paternal parent. It was observed that all One well-recognized means by which to increase the soybean gene pool is to introduce genes from wild perennial relatives, principally the subgenus Glycine. Members of this subgenus possess many identified agCommunicated by J. MacKey ronomically favorable characteristics not readily found M. H. Bodanese-Zanettini ([]) 9 M. S. Lauxen 9 S. N. C. Richter in the soybean. The principal desirable traits are resisS. Cavalli-Mohna Departamento de Gen6tica, Umversidade Federal do Rio Grande do tance to pathogens, such as viruses, bacteria, fungi and Sul, C.P. 15053, 91501-970 Porto Alegre, RS, Brazil nematodes; tolerance to stress, such as saline, drought, heat, cold; and resistance to herbicides (see Newell et al. C.-E. Lange FUNDACEP-FECOTRIGO, RS 342- Km 14, Cruz Alta, RS, Brazil 1987; Shoemaker et al. 1990; CoNe and Schapaugh 1990 for the listings of specific references). All the perennial P. J. Wang soybean relatives are wild plants of tropical origin and it Agriculture Biotechnology, Laboratory, Nauonal Chung Hsmg University Talchung, Taiwan is highly likely that these plants have evolved numerous traits specific for tropical environments that will be of C.Y Hu Biology Department, Win. Paterson College of New Jersey, Wayne, great use in the semitropical Brazilian soybean improveNJ 07470, USA ment program. Abstract Employing a different culture strategy, we

704

The difficulty in transferring genes from the subgenus Glycine to the cultivated soybean gene pool is due to recalcitrant sexual incompatibility of the post-fertilization type. The pollen tube grows normally and enters the ovule within 24 h after pollination in both reciprocal crosses of wild perennial Gl)'cine x G. max (Singh and Hymowitz 1987), but the hybrid embryos are aborted at an early developmental stage due to the degeneration of the endosperm (Palmer and Hadley 1968; Ladizinsky etal. 1979). By means of histological examinations, Sakai and Kaizuma (1985) found that some of the lateaborting hybrid pods contained embryos well into the heart stage that had the potential to develop further but at very low growth rates. The standard rescue procedure for post-fertilizationincompatible hybrids is to excise the immature embryos (or the ovules) and culture them in vitro (Hu and Zanettini 1995). The use of embryo and ovule cultures to rescue incompatible soybean intersubgeneric crosses has been investigated world-wide. Out of the 15 known species in the subgenus Gl)'cine, rescue has been reported in crosses between soybean and the only 2 known Glycine tetraploid species, G. tabacina and G. tomentella, and 2 diploid species, G. canescens and G. clandestina. Most of such hybridization attempts have been carried out in the USA (Newell and Hymowitz 1982: Newell et al. 1987; Singh and Hymowitz 1987: Singh et al. 1987; Shoemaker et al. 1990; Coble and Schapaugh 1990). Successful hybrid rescue has also been reported for the Australian (Broue et al. 1982) and Korean (Chung and Kim 1990) soybeans. Japanese workers (Sakai and Kaizuma 1985) made a large number of crosses for use in histological examinations but did not attempt embryo rescue. Our laboratory is concentrating on the improvement of Brazilian soybeans, and here we report our initial success in the wide crosses between Brazilian soybean cultivars and the subgenus Gl),cine.

and immedmtely pollinated with pollen from the neM v opened flowers of the perennial Glycme species. Only three or four buds fiom each soybean raceme were used for hybridization. To encourage the retention and growth of the pods, the hybridized gynoecIa were sprayed dail) with glbberelhc acid (100 mg...'l)for 20 days (Singh and Hymowitz 1987; Chung and Kim 1990).

Hybrid rescue Immature putative hybrid pods were harvested between 20 30 days after pollination (DAP) in 1993 and on the 20th DAP in 1994. Pods were surface-disinfected with 70% ethanol for 1 mm followed by a 20-rain soaking in 1% sodium hypochlorite with a trace amount of Tween. After being rinsed with three changes of sterile, distilled water, the pods were dissected under a stereo-microscope and the embryos were excised and cultured. The seed and embryo sizes were measured during dissection The culturing procedure was divided into the following three stages. II) embryonic development, (II) maturation and dormancy and (III) germination and seedling development. The cultures were maintained at 25: _+ 1 :C under a 16-h photoperiod of approximately 2.0 klx fluorescent light for all three stages. The newly excised embryos were placed in 15 x 40-mm glass tubes containing 2 ml of liquid Stage I Embryonic Development Medium (EDM). The EDM consisted of B5 Long medium formulanon [an organic-enriched B5 medium (Gamborg et al. 1968): Carolina Biological Supply Co, Burlington, N.C., see Hu et al. 1995 for the orgamc ingredients] supplemented with Yeung's amino acids (Yeung and Sussex 1979; 500 ing::l glutamine, 100 ing'l serine, 100 mg '1asparagine and 250 mg'l casein hydrol3'sate), 1 I,tM BAP, 0.1 btM NAA and 4% sucrose. Alter 22 (for larger embryos) to 61 (for smaller embryos) days, the embryos were transferred to Stage II Maturation and Dormancy Medium (MDM) atld incubated under reduced hght. The M D M consisted of B5 Lotlg medium with 10% sucrose, 0.5~ activated charcoal and 1% agar. After 31-56 days incubation, the mature, dormant embryos were transferred onto Stage III Germination and Seedhng Development Medium (GSM) consisting of SH medium (Schenk and Hildebrandt 1972) ~ith 1% sucrose and 0.6% agar. After 13-67 days, plantlets with well-developed root systems were transplanted to 250-ml plastic pots containing a 3'1 mixture of peat and carbonized rice hulls.

Hybrid identification

Materials and methods Plant material Plant material used in this study were soybean strains fiom the subgenus Sq]a (Moench) F. J. Herin. and wild perennial species from the subgenus Gl,rcme Wllld. The soybeans (Glycme max, L. Merr; 2n = 2x = 40) ~ere Brazilian culfivars and breeding lines CEP-10, CEP-12-Cambarfi, CEP-20-Guajuvira. CEP-26-Umbu, CEP-7403. COBB, IAS-5. PRATA and RS-7-Jacui The wild perennial species were G canescens F. J. Herm. 12n = 2x = 40: G 2528. G 9937 and PI 440151). G. microph)'Ila (2n = 2x = 40; PI 509488), G. tabacina (Labill.) Benth. (2n = 4 x = 80, PI 339661. PI 505197, PI 509495. PI 509496 and PI 509498) and G. tomentella Hayata (2n = 4x = 78: G 9941, G 9943 and PI 5095011. Seeds of wild perennial species were obtained from CSIRO, Canberra, Australia. and The Asian Vegetable Research and Development Center-AVRDC, Shanhua, Talnan, Republic of China Soybeans and wild perennials used in the crossing program were grown in a temperature-regulated (27 ~ _+ 2 ~C) greenhouse at FUNDACEP-FP_COTRIGO, Cruz Alta, RS, Brazil.

flOSSeS

Crosses were made in 1993 and 1994 with soybeans as the female parents. The young buds were emasculated 2 or 3 days before anthesis

Root-tip mitosis, mlcrospore meiosis and leaf lsoenz)me analysis were used to identl~.' the hybrid status of the resultant plants. Root tips were harvested from the putative hybrid plantlets before they were transplanted ex vitro. Root-tip squash and chromosome couuts were carried out according to Palmer and Heer (1973). Floial buds which were harvested from greenhouse-grown hybrid plants were fixed (absolute ethanol: acetic acid = 3'1) for 24 h and stored in 70% ethanol. Anthers were squashed in 0.6~ propiomc carmine for analysis of the meiotic chromosome behavior m the pollen mother cells. Leaves from 2 greenhouse-grown adult putative hybrid plants were assayed for their isoenzylne patterns. Isoenzyme analysis of glutamate oxalacetate transammase (GOT). peroxmdases (PER), superoxide dismutases (SOD), malate dehydrogenases (MDH). amylases (AMY) and esterases (EST) were carried out by horizontal polyacrylamlde gel electrophoresis. The migration conditions utilized for each enzyme system were: 7% gel and Brown's 11983) buffers for GOT; 7% gel and Scandalios" (1969) buffers for PER, SOD and AMY; 6~ gel and Roose and Gottlieb's 11976) buffers for MDH: 8~ gel and Scandalios" 11969) buffers for EST. The gels were run at 10 V:cm and stained as descr%ed by the following authors: Vallejos 11983) with modifications for GOT: Gottlleb {1973) for PER; Brewer ~1970) for MDH; Scandahos (1969) for EST: Chao and Scandahos 11972) with modifications for AMY. SOD was stained using the methods of Brewer (1970) for glutamate deh3drogenase and the staining Imxture incubated under illumination.

705 100

Grafting and chromosome doubling

"

~

a

90

To increase the plant number, we cloned the confirmed hybrid plants using the grafting procedure described by Newell and Hymowitz (1979). Buds from both the original and the grafted clonal hybrid plants were treated with colchicine (0.1% or 0.2 %) to induce chromosome doubling as described by Cheng and Hadley' (1983).

80 70 tO

{.,-

60 50 40

-o 30 O ca_ 20

Results

Crosses and hybrid rescue

10

A

0

Five perennial Glycine accessions comprising three species, and seven lines of G. max were used as parent lines in 1993 (Table 1A). From approximately 400 crosses, 19 putative hybrid pods were harvested and 23 embryos cultured. Eleven embryos survived the initial Stage I culture, developing to the plantlet stage and successfully transplanted ex vitro. Two of these hybrid plants, both with G. tomentella G 9943 as the paternal parent, survived to maturity. The maternal parents were soybean strains CEP-12 and CEP-7403. Nine wild perennial Glycine accessions representing four species and four Brazilian soybean cultivars were used as parent lines in 1994 (Table 1B). The pod retention rates up to 20 DAP are presented in Fig. 1. Nearly identical trends on the retention curves were observed regardless of parental genotypes. In general, the pod retention rates dropped quickly and steadily immediately after pollination until the 8th day and stabilized at approximately 10% up to 20 DAP ( the pod harvesting day). From 678 crosses, 62 putative hybrid pods were harvested.

J i l l

I

I

I

I

I

I

I

l

l

l

i

l

l

l

100

o.d

90

+

o~ 8070

e

xf

I

60 c

o_ 50E 40-o 3 0 0

a. 20100

a

I

I

I

i

L

I

I

I

I

I

[

I

[

I

I

J

I

I --'

4 5 6 7 8 9 1011 121314151617181920 Days after polhnatlon

0 123

Fig. IA, B Pod retention rates up to 20 days after 1994 crosses betv~een four Brazilian soybean stratus and nine perenmal accessions representing four Glycine species. A Rates based on the maternal parents, B rates based on thepaternal parents a CEP-12 x G. spp., b CEP-26 • G. spp., c IAS-5 x G. spp., d G. max x G tomentella. e G. m ax x G. t a b a c i , a . f G , m ax x G. spp.

=

Table 1 Hybrid embryo rescue m intersubgeneric crosses between Brazilian soybean (G. max) cultix.ars and ~ ild perenmal relaUves (G. ca~2 G. canescens, G mic G. mlcrophylla, G. tab G. rabacma, G. urn7 G. tomentella)

A) 1993 crosses Combinatxon

Pod set

Number of embryos or plants Stage 1

G. G. G. G.

ma x max max max

(2)a x G. can 12lb (3) x G t a b ( l ) (1) x G. tom AVRDC-G9941 (2) x G. tom AVRDC-G9943

G m a x x G. spp.

Stage 2

Stage 3

Ex ritro

Adult

3 3 2 II

3 5 2 13

2 1 2 6

2 1 2 6

2 l 2 6

0 0 0 "

19

23

11

11

1i

2

B) 1994 crosses Combination

Number of crosses

Pod Set

Seed ~ T

G. G G. G G. G

ma x (l) a X G. cart (1) b m a x (1) x G mic{1) max (41 x G. tab (4) m a x t 4 ) x G. tom AVRDC-G9941 m a x 13) x G. tom AVRDC-G9943 m a x (4'} x G. tom PI509501

G. max • G. spp.

Number of embryos or plants SG2

Stage 1

Stage 2

Stage 3

E x t'itro

GH a

Adult

16 13 211 115 208 115

1 1 16 7 30 7

0 1 32 16 61 14

0 0 16 7 33 11

0 0 16 7 33 11

0 0 10 5 28 11

0 0 10 5 28 11

0 0 1 1 12 8

0 0 0 0 5 2

0 0 0 0 5 0

678

62

124

67

67

54

54

22

7

5

Number of soybean cultivars used in the crosses b Number of wild perennial accessions used m the crosses

c Seed: T, total; SG2. seed size group 2 a GH, Plants transferred to the greenhouse

706 Three distinct seed-size groups were observed from the 124 immature seeds obtained in 1994. The 52 Group 1 seeds were less than 1.3 mm in length, and we were not confident that the minute structures dissected from them were in fact embryos because none of them showed embryonic growth in culture. It is highly probable that most of them were empty seed coats. Therefore, we excluded this seed group from our statistics. The 67 Group 2 seeds ranged in sizes from 1.9 to 5.0 mm (average length = 3.05 ram) with all containing small, shrunken embryos. Group 3 consisted of 5 seeds larger than 5.0 mm containing apparently healthy embryos, probably resulting from self-pollination. Sixty-seven embryos were dissected and cultured from Group 2 seeds; 54 survived the initial culture and progressed through the maturation and germination stages with 22 of these forming plantlets that were transplanted ex vitro. All but 1 of these plantlets had G. tomentella as the paternal parent. It appeared that there was no correlation between the initial seed size and the embryo survival rate among the surviving embryos excised from the Group 2 seeds. Seven plantlets, with G. ~omentella as the paternal parents, outlived ex vitro shock and were transplanted into a greenhouse. Of these 7, 5, all with accession G 9943 as the paternal parent, reached maturity. The initial size of these 5 was: 1.9, 2.8, 3.0, 3.5 and 4.2 mm. The maternal parents of these 5 mature plants included three soybean cultivars.

Five embryos were excised from Group 3 seeds and cultured; all them produced plantlets that were transferred ex citro.

Hybrid identification Both mature plants from 1993 crosses and all 22 ex vitro plantlets derived from Group 2 medium-sized seeds of the 1994 crosses expressed hybrid morphology as described by Newell and Hymowitz (1982). All 5 plantlets derived from Group 3 large seeds showed maternal soybean phenotypes. Root mitotic metaphase chromosome examination confirmed the hybrid status of both of the mature 1993 plants. From the 22 Group 2 plantlets (1994) we attempted chromosome analysis on 15, of which 3 gave unreadable chromosome spreads. All of the 12 analyzed plantlets were hybrids with chromosome numbers 2n = 59 (Fig. 2a). We were not able to count the chromosomes from 2 plantlets derived from Group 3 seeds with the other three given their chromosome numbers of

Fig. 2a-d Cytologicalfiguresin F 1(a-c) and colchicine-doubled(d) mterspecific hybrids between Brazilian soybean cultlvars (2n = 40) and G. tomentella G9943 12n = 78). a Mitotic metaphase chromosomes(2n = 59),b diakinesis(2n = 59)with 551 + 2 II, c metaphaseI (2n = 59) showing 33 I + 13 1I, d dlakinesls(2n = 118/showing 16 I+5111 Bar:10~tm

707

2n = 40. On the basis of their phenotypes and chromosome counts, they must be the selfing products of the maternal soybean parents. G. max and G. tomemella parents showed different isoenzyme patterns and specific bands in all of the systems analyzed. Hybrid status was confirmed by the presence of additive isoenzyme patterns in the putative hybrid plants (Fig. 3). For GOT, two isoenzymes were specific for the two species: the band o f R M = 1.17 only occurred in G. max (m) and that o f R M = 1.00, only in G. wmentella (t). The hybrids (h) presented both bands as well as a hybrid band with intermediate mobility (RM = 1.09), which could be interpreted as an heteromer formed by the polypeptide subunits fiom bands 1.17 and 1.00. For PER, G. max showed an isoenzyme with RM = 1.32, and G. tomemella presented another band

Fig. 3 Isoenzyme patterns. Gels showing peroxidase (PER) and glutamate oxalacetate transaminase (GOT) patterns of G. max (m), G. tomemella (t) and their hybrids (h)

with RM = 1.00, with the hybrids presenting both bands. All other systems showed similar profiles, thereby confirming their hybrid status. Chromosome doubling Colchicine (0.1% or 0.2%) treatments were performed on 5 mature hybrid plants and their grafted clones. The 5 hybrid plants (with the maternal origin and number of clonal plants treated in parenthesis) were: Hyb 5 (CEP-

Table 2 Average chromosome configurations m F 1 and colchlcme doubled mterspecific hybrids between Brazilian soybean (2n= 401 cultlvars (female) • G tomemella G9943 (2n = 78) (male) Soybean cultlvars used as female

Total PMCs

Chromosome configurauons 2n

CEP 12 (2)~

118

59

CEP 26 {1)

40

59

10

118

CEP-7403 ~1)

30

59

I A S - 5 (2)

41

59

a Number of plants from separate fertilization events

I

II

III

IV

43.30 (27 571 44.40 (33 55/ 20.70 (4-42) 43.i7 (33-55) 44.49 129 53)

7.85 (1-17) 7.30 (2 13) 46.90 (37-57) 7.90 (2-13t 7 24 (3 15)

0 90 (0-4)

0.20 (0-1) 0.06 ~0 1)

0.03 (0 1)

708

12; 3), Hyb 20 (CEP-7403; 8), H 3 2A (IAS-5: 7). H 5 3B (CEP-12: 5) and H 9 3B (CEP-26: 4). One of the four clonal plants of H 9 3B showed chromosome doubling having 118 chromosomes at meiosis (Table 2). Meiotic chromosomal behaviour of the hybrids Pollen mother cell meiotic chromosomal configurations of the confirmed hybrids are shown in Table 2 and Fig. 2. The overall average of univalents, bivalents. trivalents and quadrivalents for the 6 F~ hybrids analyzed was 43.71, 7.64, 0.01 and 0.01, respectively. The same values for the colchicine-doubled hybrid were 20.7, 46.9, 0.9 and 0.2, respectively.

Discussion

rescuing percentage for soybean with our culture strategy. All of the 7 mature plants possessed the expected hybrid chromosome number of 59. The meiotic chromosome configurations observed by us (Table 2) were similar to those recorded by Newell and Hymowitz (1982) but they had on average higher numbers of bivalents than those reported by Newell et al. (1987) on G. max x G. tomemella crosses. Although a portion of the bivalents and multivalents might result from chromosome interactions within the G. romemella (n -- 2X) or soybean (n = X; Crane et al. 1982) genomes, some of them should result from chromosome interactions between the two genomes and provide an opportunity for genomic exchanges. We have obtained 1 chromosome-doubled plant from our F~ hybrid clones. Newell et al. (1987) obtained chromosome doubling in 6 of their 7 G. max x G. tomentella F~ hybrid clones. Since the F~ hybrids are perennial in nature and vigorous in growth, an unlimited number of clonal plants via grafting can be obtained from each F 1 genotype. Fertile progenies can be expected to be obtained from them after a sufficient number of colchicine treatments are performed. Shoemaker et al. (1990) obtained fertile F 2 and F 3 plants with exclusively a 2n = 40 soybean genome from 1 of the colchicinedoubled Ft hybrids. An analysis ofisoenzymes indicated that regions of the G. tomemella genome were retained. This fact suggests that the complete G. ~omemella chrolnosome complement from the F~ hybrid might automatically be eliminated after genetic exchange had taken place. Preferential elimination of one parental genome is common in human-mouse somatic hybrid cells (Cowell 1992). It is also well-documented in crosses between Hordeum L,ulgare and H. bulbosum (Kasha and Kao 1970) and Triticum aestivum with Hordeum bulbosum (Barclay 1975). All 7 F~ hybrids were obtained using G. tomentella G 9943 as the sole paternal parent. It appears that all of the Brazilian soybean strains that we tested were capable of producing mature F~ plants when crossed with this accession. We observed no correlation between the initial seed size at dissection and the hybrid survival rate, but the genotype and the physiological condition of the excised embryo probably play roles in determining rescue success. Figure 1 shows that hybrid pod-retaining rates dropped quickly to about 10% during the first 8 days, similar to that observed by Chung and Kim (1990). Although the pod number stayed largely unchanged from the 9th to the 20th day, the physiological condition of the embryos would be expected to deteriorate during this period. Therefore, it would be advantageous to develop strategies to rescue embryos prior to 8 DAP. Obviously, the path ahead of us is to obtain F~ plants from a wider range of accessions and species of the perennial Glycine.

Because hybrid embryos between soybean and perennial Glycine are feeble, poorly developed and at very early developmental stages when excised, the success rate for in vitro rescuing is usually low. From 1,534 hybrid embryos, Newell et al. (1987~ obtained 22 tl.4%) in vitro-germinated plantlets; 16 (1.04%) of these survived ex vitro transfer, greenhouse conditions and reached maturity. Chung and Kim (1990) rescued 45 hybrid embryos with 17.8% reaching in vitro germination. Hovever, the authors did not present data on the number of plants that reached maturity. Their higher success rate was partially due to the fact that only G. tomentella was used in their crosses. This species appears to be more cross-compatible with soybean than the other Glycine species (Newell et al. 1987). The highest rate of hybrid embryo rescue has been reported by CoNe and Schapaugh (1990). From six excised hybrid embryos from a G. max x G. tomentella cross, four in vitro plantlets were obtained on B5 salts with Williams' (1978) vitamins and 3% sucrose. Again, no mention was made of the number of plants that survived ex vitro transplanting. In order to improve the rate of rescuing, we have developed a different culturing strategy. In addition to using a very rich medium to nurse the newly dissected embryos, we exposed the in vitro-growing embryos to a maturation and dormancy period using a high osmotic medium (Ranch et al. 1985). It is known that when soybean embryos are partially dehydrated and dormant. a higher percentage of them will germinate in vitro (Buchheim et al. 1989). With our protocol, from 23 rescued putative hybrid embryos in 1993, we obtained 1l (47.8%) in vitro-germinated plantlets and 2 (8.7%) mature hybrid plants. In 1994, from 67 putative hybrid Group 2 seeds we obtained 22 132.8%) in vitro-regenerated plantlets and 5 (7.5%) greenhouse-matured plants. The pooled data, including the putative hybrid embryos obtained in 1993 and 1994, gives an average of 36.7% regenerated plantlets and 7.8% adult plants. Acknowledgements This work was supported in part by grants from Clearly we have greatly improved the hybrid embryo F A P E R G S and FINEP. The initml development of the soybean

709 embryo culture was supported by a supplemental summer grant to I.M. Sussex from NSF. The visiting research professor. C.Y. Hu, was supported by grants from RHAE.'CNPq in 1990 and 1992 and from FAPFRGS in 1995.

References Abdelnoor RV, de Barros EG, Moreira MA (1995) Determination of genetic diversity within Brazilian soybean germplasm using random amplified polymorphic DNA techniques and comparative analysis with pedigree data. Braz J Genet 18:265 273 Barclay I 11975~ High frequencies of haploid production in wheat (Triticum aestttum) by chromosome elimination. Nature 256: 410-411 Brewer GJ (19701 An introduction to isozyme techniques. Academic Press. New York London Broue P, Douglas J, Grace JP, Marshall DR (1982) Interspecific hybridlzatxon of soybeans and perennial Glycule species indigenous to Austraha via embryo culture. Euphytica 31:715 724 Brown AHD (1983) Barley. In Tanksley SD, Orton TJ (eds) Isozymes in plant genetics and breeding, part B. Elsevier Science Pub, Amsterdam, pp 57 77 Buchheim, JA, Colburn SM, Ranch JP (19891 Maturation of soybean somatm embryos and the transition to plantlet growth. Plant Physiol 89:768-775 Chao SE, Scandalios JG (1972) Developmentally dependent expression of txssue speofic amylases in maize. Molec Gen Genetics 115:1-9 Cheng SH, Hadley' HH (19831 Studies in polyplotdy in soybeans:a simple and effective colchicine technique of chromosome doubling for soybean IGlycme max tL.) Merr.) and its wild relatives. Soybean Genet Newsl 10:23-24 Chung GH, Klm JH 11990) Production of interspeclfiC hybrids between Glycme max and G tomentella through embryo culture Euphytica 48'97 110 Coble C J, Schapaugh WT Jr (1990} Nutrient culture medium components affecting plant recovery from Immature embryos of three Glycine genotypes and an mterspecific hybrid grown in ritro. Euphytlca 50:127-133 Coweli JK (19921 Somatic cell hybrids in the analys~s of the human genome. In:Rooney DE, Czepulkowskl BH (eds) Human cytogenencs. A pracncal approach, vol. 1 :malignancy and acquired abnormalities. Oxford Umversity Press, Oxford New York Tokyo. pp 235 252 Crane CF, Beversdorf WD, Bingham ET( 1982t Chromosome pairing and associations at meiosis m haploid soybean (Gl)'clne max). Can J Genet Cytol 24:293-300 Gamborg OL, Miller RA, Ofima K t1968) Nutrient requirements of suspension cultures of soybean root cells. Exp Cell Res 50: 151 158 Gottlieb LD (1973) Genetic differentiation, sympatric specianon, and the origin of a diploid species of Stephanomeria. Amer J Bot 60' 545-553

Hiromoto DM, Vello NA (1986) The geneuc base of Brazihan soybean (Glycine max (L.I Merrill) cultivars Braz J Genet 9: 295-306 Hu CY, Zanettini MH 119951 Embryo culture:embryo rescue for wide cross hybrids. In:Gamborg OL, Philhps GC (eds/Textbooks for training, education and research in applied and basic plant biotechnology, vol 1 :fundamental methods of plant cell, tissue, and organ culture and laboratory operations. Springer, Berlin Heidelberg New York, pp 129 141 I-tu CY. Yln GC, Zanettini MH (in press) Haploid of soybean, a rex.iew article. In :Jain SM, Sopory, Veilleux reds) In vitro haploid production In higher plants. Kluwer Acad. Dordrecht. The Netherlands Kasha K J, Kao KN ~1970) High frequency of haploid production in barley (Hordeum ruhjare L.). Nature 225. 874-876 Ladizinsky G. Newell CA, Hymowitz T (1979) Wide crosses in soybeans.prospects and limitations. Euphytlca 28:421 423 Newell CA, Hymowitz T 11979) Flower induction in Glycme tomentella following grafting onto G. tnax (L.) Merr. Crop Set 19 121-123 Newell CA, Hymowilz T 1,1982) Successfnl wxde hybridization between the soybean and a ~fld perennial relative G. tomemella Hayata. Crop Sci 22:1062-1065 Newell CA, Delannay X, Edge ME (1987) Interspecific hybrids between the soybean and wild perennial relatives. J Hered 78 : 301-306 Palmer RG, Hadley HH (1968) Interspecific hybridization in Glycine, subgenus Leptoc) amus. Crop Sci 8:557-563 Palmer RG, Herr H (1973} A root tip squash technique for soybean chromosomes. Crop Sci 13:389 391 Ranch JP. Oglesby L, Zielinskl AC (1985) Plant regeneration from embryo-derivedtissue cultures of soybeans. In Vitro Cell Dev Bml 21 : 653-658 Roose ML, Gottlieb LD (1976) Genetic and biochemical consequences of polyplo~dy in Tragopogon. Evolution 30 : 818-830 Sakai T, Kaizuma N (1985) Hybrxd embryo formation in an mtersubgeneric cross of soybean (Glycine max Merill) with a wdd relam'e (G. tomemella Hayata). Jpn J Breed 35:363 374 Scandalios JG (1969) Genetic control of multiple molecular forms of enzymes in plants:a review. Biochem Genet 3:37-79 Schenk RU, Hildebrandt AC (1972) Medium and techmques for induction and growth of monocotyledonous and dicotyledonous plant cell cultures. Can J Bot 50:199 204 Shoemaker RC, Heath MS, Skorupksa H, Delanna> X. Edge M, Newell CA (1990) Fertile progeny of a hybridization between soybean [Glyclne max (L.) Merr.] and G. tomentella Hayata. Theor Appl Genet 80:17 23 Singh RJ. Hymowitz T (1987) Intersubgenerlc crossablhty in the genus Glycine Wil/d. Plant Breed 98:171-173 Singh RJ, Kollipara KP. Hymowitz T (1987) lntersubgeneric hybridization of soybeans with a wild perennial species. Glycme clandestina Wendl. Theor Appl Genet 74'391 396 Vallejos E (1983) Enzyme activity staining In:Tanksley SD, Orton TJ (eds) Isozylnes and plant genetics and breeding, part A. Elsevier Science Pub, Amsterdam, pp 469 516 Yeung EC, Sussex IM (1979) Embryogeny ofPhaseolus cocclneus: the suspensor and the growth of the embryo-proper in vitro. Z Pflanzenphysiol 91:423-433